EXPERIMENTAL DESIGN
Title
Evaluation of a Diagnostic Test to Identify Classic Story: The Application of the Scientific Method in the Study of Modern Story Structure
Purpose
To evaluate The Story Template as a post-production diagnostic tool to identify and differentiate between “classic” and “non-classic” stories
Hypothesis
The Story Template, when applied as a diagnostic test, is a sensitive, specific, and accurate system that can identify “classic” versus “non-classic” literature using a set criteria including story structure as its main component.
Test Group: The Story Template algorithm
Based on The Story Template, The Story Template algorithm classifies stories as “classic” or “non-classic” according to each subjects score using its list of criteria (16 attributes of story described in The Story Template) for “classic” versus “non-classic”.
Control Group: The Gold Standard
The Gold Standard classifies stories as “classic” or “non-classic” based on the components of timelessness (required: original form created at least 25 years ago=1985) and impact (required: work must have been reproduced in one or more forms, one of which must have been a movie).
PROCEDURES
Subject Selection
A group of stories were chosen randomly after they met the following criteria:
- Story must be available in full-length movie format (excluding TV shows).
- Original form must have been published/released prior to 1985 (requirement excluded by the prediction test group).
Gold Standard Classification
1. All subjects are then divided into two categories: “classic” and “non-classic”.
2. If the subject was reproduced at least one time after the original form, then the work is considered “classic”. Though the financial investment needed to redo or supplement the original work does not necessarily have the same quality, it demonstrates an interest in the work for the expected profit it would have, because the original story already has an established good reputation. The following scenarios are examples of a work’s reproduction:
a. A book adapted into full-length movie form
b. A movie adapted into a TV series
c. A movie remade
d. A sequel or prequel written or filmed
3. If the subject was not reproduced in any form since the original, then the work is considered a “non-classic”, because there was no tangible interest shown in the work despite its opportunity (time requirement equal to “classics” in which the “non-classic” works still did not perform).
The Story Template Algorithm Classification
1. Construct scene-by-scene story summary with timestamps shown below:
1. Start time of scene
2. Duration of scene
3. Scene’s percent of whole story according to time ([scene duration/movie duration] * 100)
4. Percent of story at which scene starts
2. Identify story posts and other major structural components shown below:
1. [proprietary]
2. [proprietary]
3. [proprietary]
4. [proprietary]
5. [proprietary]
6. [proprietary]
7. [proprietary]
i. [proprietary]
ii. [proprietary]
iii. [proprietary]
8. [proprietary]
9. [proprietary]
10. [proprietary]
11. [proprietary]
12. [proprietary]
13. [proprietary]
i. [proprietary]
ii. [proprietary]
iii. [proprietary]
14. [proprietary]
3. **Optional: Identify character strands listed below to help in identifying story posts and other major structural components listed above:
1. “A-Strand”: [proprietary]
2. “B-Strand”: [proprietary]
3. “C-Strand”: [proprietary]
4. “D-Strand”: [proprietary]
5. “E-Strand”: [proprietary]
4. Compare the previously identified structural components for each subject with The Story Template’s ideal duration and location for each of those components all of which are listed below.
Gold Standard: Components’ Ideal Duration and Location
Component |
Duration |
Location (Start) |
|
[proprietary] |
[proprietary] |
[proprietary] |
|
[proprietary] |
[proprietary] |
[proprietary] |
|
[proprietary] |
[proprietary] |
[proprietary] |
|
[proprietary] |
[proprietary] |
[proprietary] |
|
[proprietary] |
[proprietary] |
[proprietary] |
|
[proprietary] |
[proprietary] |
[proprietary] |
|
[proprietary] |
[proprietary] |
[proprietary] |
|
[proprietary] |
[proprietary] |
[proprietary] |
|
[proprietary] |
[proprietary] |
[proprietary] |
|
[proprietary] |
[proprietary] |
[proprietary] |
|
[proprietary] |
[proprietary] |
[proprietary] |
|
[proprietary] |
[proprietary] |
[proprietary] |
|
[proprietary] |
[proprietary] |
[proprietary] |
|
[proprietary] |
[proprietary] |
[proprietary] |
|
[proprietary] |
[proprietary] |
[proprietary] |
|
[proprietary] |
[proprietary] |
[proprietary] |
5. Calculate percent error for duration and location of each component.
6. Average all percent errors for all components in two categories: duration and location to get two final percent error scores.
7. If both these scores are equal to or less than 5%, then the subject is identified as a “classic”. If either of these scores is greater than 5% and the collective percent error from both duration and location is less than 10%, then the subject is still identified as “classic”. If either of these scores is greater than 5% and the collective percent error from bother duration and location is greater than 10%, then the subject is identified as a “non-classic”.
8. If one of the sixteen above components simply does not exist in the work, then the work is considered a non-classic.
Comparison of Test versus Control Group Classifications
1. The control group Gold Standard broke all subjects into the categories of “classic” and “non-classic” in an above stage.
2. If the test group Story Template Algorithm identifies “classic” subjects correctly (according to the Gold Standard’s classification), then this will contribute to a high sensitivity score.
3. If the test group Story Template Algorithm identifies “non-classic” subjects correctly (according to the Gold Standard’s classification), then this will contribute to a high specificity score.
4. Accuracy was calculated using both the “classic” and “non-classic” components.
|
|
GOLD STANDARD |
|
|
Classic |
Non-Classic |
|
|
Algorithm Identifies Classic |
a |
b |
|
Algorithm Identifies Non-Classic |
c |
d |
5. Percent Sensitivity = [a / (a + c)]*100
6. Percent Specificity = [d / (b + d)]*100
7. Percent Accuracy = [(a + d) / (a +d) + (b +c)]*100